Search

Molecular imaging is established as an indispensable tool in various areas of cancer research, ranging from basic cancer biology and preclinical research to clinical trials and medical practice. In particular, the field of fluorescence imaging has experienced exceptional progress during the last three decades with the development of various in vivo technologies. Within this field, fluorescence microscopy is primarily of experimental use since it is especially qualified for addressing the fundamental questions of molecular oncology. As stimulated emission depletion (STED) nanoscopy combines the highest spatial and temporal resolutions with live specimen compatibility, it is best-suited for real-time investigations of the differences in the molecular machineries of malignant and normal cells to eventually translate the acquired knowledge into increased diagnostic and therapeutic efficacy. This thesis presents the application of STED nanoscopy to two acute topics in cancer research of direct or indirect clinical interest. The first project has investigated the structure of telomeres, the ends of the linear eukaryotic chromosomes, in intact human cells at the nanoscale. To protect genome integrity, a telomere can mask the chromosome end by folding back and sequestering its single-stranded 3’-overhang in an upstream part of the double-stranded DNA repeat region. The formed t-loop structure has so far only been visualized by electron microscopy and fluorescence nanoscopy with cross-linked mammalian telomeric DNA after disruption of cell nuclei and spreading. For the first time, this work demonstrates the existence of t-loops within their endogenous nuclear environment in intact human cells. The identification of further telomere conformations has laid the groundwork for distinguishing cancerous cells that use different telomere maintenance mechanisms based on their individual telomere populations by a combined STED nanoscopy and deep learning approach. The population difference was essentially attributed to the promyelocytic leukemia (PML) protein that significantly perturbs the organization of a subpopulation of telomeres towards an open conformation in cancer cells that employ a telomerase-independent, alternative telomere lengthening mechanism. Elucidating the nanoscale topology of telomeres and associated proteins within the nucleus has provided new insight into telomere structure-function relationships relevant for understanding the deregulation of telomere maintenance in cancer cells. After understanding the molecular foundations, this newly gained knowledge can be exploited to develop novel or refined diagnostic and treatment strategies. The second project has characterized the intracellular distribution of recently developed prostate cancer tracers. These novel prostate-specific membrane antigen (PSMA) inhibitors have revolutionized the treatment regimen of prostate cancer by enabling targeted imaging and therapy approaches. However, the exact internalization mechanism and the subcellular fate of these tracers have remained elusive. By combining STED nanoscopy with a newly developed non-standard live cell staining protocol, this work confirmed cell surface clustering of the targeted membrane antigen upon PSMA inhibitor binding, subsequent clathrin-dependent endocytosis and endosomal trafficking of the antigen-inhibitor complex. PSMA inhibitors accumulate in prostate cancer cells at clinically relevant time points, but strikingly and in contrast to the targeted antigen itself, they eventually distribute homogenously in the cytosol. This project has revealed the subcellular fate of PSMA/PSMA inhibitor complexes for the first time and provides crucial knowledge for the future application of these tracers including the development of new strategies in the field of prostate cancer diagnostics and therapeutics. Relying on the photostability and biocompatibility of the applied fluorophores, the performance of live cell STED nanoscopy in the field of cancer research is boosted by the development of improved fluorophores. The third project in this thesis introduces a biocompatible, small molecule near-infrared dye suitable for live cell STED imaging. By the application of a halogen dance rearrangement, a dihalogenated fluorinatable pyridinyl rhodamine could be synthesized at high yield. The option of subsequent radiolabeling combined with excellent optical properties and a non-toxic profile renders this dye an appropriate candidate for medical and bioimaging applications. Providing an intrinsic and highly specific mitochondrial targeting ability, the radiolabeled analogue is suggested as a vehicle for multimodal (positron emission tomography and optical imaging) medical imaging of mitochondria for cancer diagnosis and therapeutic approaches in patients and biopsy tissue. The absence of cytotoxicity is not only a crucial prerequisite for clinically used fluorophores. To guarantee the generation of meaningful data mirroring biological reality, the absence of cytotoxicity is likewise a decisive property of dyes applied in live cell STED nanoscopy. The fourth project in this thesis proposes a universal approach for cytotoxicity testing based on characterizing the influence of the compound of interest on the proliferation behavior of human cell lines using digital holographic cytometry. By applying this approach to recently developed live cell STED compatible dyes, pronounced cytotoxic effects could be excluded. Looking more closely, some of the tested dyes slightly altered cell proliferation, so this project provides guidance on the right choice of dye for the least invasive live cell STED experiments. Ultimately, live cell STED data should be exploited to extract as much biological information as possible. However, some information might be partially hidden by image degradation due the dynamics of living samples and the deliberate choice of rather conservative imaging parameters in order to preserve sample viability. The fifth project in this thesis presents a novel image restoration method in a Bayesian framework that simultaneously performs deconvolution, denoising as well as super-resolution, to restore images suffering from noise with mixed Poisson-Gaussian statistics. Established deconvolution or denoising methods that consider only one type of noise generally do not perform well on images degraded significantly by mixed noise. The newly introduced method was validated with live cell STED telomere data proving that the method can compete with state-of-the-art approaches. Taken together, this thesis demonstrates the value of an integrated approach for STED nanoscopy imaging studies. A coordinated workflow including sample preparation, image acquisition and data analysis provided a reliable platform for deriving meaningful conclusions for current questions in the field of cancer research. Moreover, this thesis emphasizes the strength of iteratively adapting the individual components in the operational chain and it particularly points towards those components that, if further improved, optimize the significance of the final results rendering live cell STED nanoscopy even more powerful.